ICU-20441 force Gannen use for ja@calendar=japanese non-numeric formats if no override (#474)
diff --git a/icu4c/source/i18n/smpdtfmt.cpp b/icu4c/source/i18n/smpdtfmt.cpp
index b691944..414dfff 100644
--- a/icu4c/source/i18n/smpdtfmt.cpp
+++ b/icu4c/source/i18n/smpdtfmt.cpp
@@ -856,6 +856,17 @@
{
if (U_FAILURE(status)) return;
+ parsePattern(); // Need this before initNumberFormatters(), to set fHasHanYearChar
+
+ // Simple-minded hack to force Gannen year numbering for ja@calendar=japanese
+ // if format is non-numeric (includes 年) and fDateOverride is not already specified.
+ // This does not update if applyPattern subsequently changes the pattern type.
+ if (fDateOverride.isBogus() && fHasHanYearChar &&
+ fCalendar != nullptr && uprv_strcmp(fCalendar->getType(),"japanese") == 0 &&
+ uprv_strcmp(fLocale.getLanguage(),"ja") == 0) {
+ fDateOverride.setTo(u"y=jpanyear", -1);
+ }
+
// We don't need to check that the row count is >= 1, since all 2d arrays have at
// least one row
fNumberFormat = NumberFormat::createInstance(locale, status);
@@ -872,8 +883,6 @@
{
status = U_MISSING_RESOURCE_ERROR;
}
-
- parsePattern();
}
/* Initialize the fields we use to disambiguate ambiguous years. Separate
@@ -4216,6 +4225,9 @@
if (ch == QUOTE) {
inQuote = !inQuote;
}
+ if (ch == 0x5E74) { // don't care whether this is inside quotes
+ fHasHanYearChar = TRUE;
+ }
if (!inQuote) {
if (ch == 0x6D) { // 0x6D == 'm'
fHasMinute = TRUE;
diff --git a/icu4c/source/i18n/unicode/smpdtfmt.h b/icu4c/source/i18n/unicode/smpdtfmt.h
index f1e7c96..a015c5be 100644
--- a/icu4c/source/i18n/unicode/smpdtfmt.h
+++ b/icu4c/source/i18n/unicode/smpdtfmt.h
@@ -1599,6 +1599,7 @@
UBool fHasMinute;
UBool fHasSecond;
+ UBool fHasHanYearChar; // pattern contains the Han year character \u5E74
/**
* Sets fHasMinutes and fHasSeconds.
diff --git a/icu4c/source/test/cintltst/cdattst.c b/icu4c/source/test/cintltst/cdattst.c
index 72dbb94..d5e3ba2 100644
--- a/icu4c/source/test/cintltst/cdattst.c
+++ b/icu4c/source/test/cintltst/cdattst.c
@@ -42,6 +42,7 @@
static void TestCalendarDateParse(void);
static void TestParseErrorReturnValue(void);
static void TestFormatForFields(void);
+static void TestForceGannenNumbering(void);
void addDateForTest(TestNode** root);
@@ -61,6 +62,7 @@
TESTCASE(TestOverrideNumberFormat);
TESTCASE(TestParseErrorReturnValue);
TESTCASE(TestFormatForFields);
+ TESTCASE(TestForceGannenNumbering);
}
/* Testing the DateFormat API */
static void TestDateFormat()
@@ -1867,4 +1869,40 @@
}
}
+static void TestForceGannenNumbering(void) {
+ UErrorCode status;
+ const char* locID = "ja_JP@calendar=japanese";
+ UDate refDate = 600336000000.0; // 1989 Jan 9 Monday = Heisei 1
+ const UChar* testSkeleton = u"yMMMd";
+
+ // Test Gannen year forcing
+ status = U_ZERO_ERROR;
+ UDateTimePatternGenerator* dtpgen = udatpg_open(locID, &status);
+ if (U_FAILURE(status)) {
+ log_data_err("Fail in udatpg_open locale %s: %s", locID, u_errorName(status));
+ } else {
+ UChar pattern[kUbufMax];
+ int32_t patlen = udatpg_getBestPattern(dtpgen, testSkeleton, -1, pattern, kUbufMax, &status);
+ if (U_FAILURE(status)) {
+ log_data_err("Fail in udatpg_getBestPattern locale %s: %s", locID, u_errorName(status));
+ } else {
+ UDateFormat *testFmt = udat_open(UDAT_PATTERN, UDAT_PATTERN, locID, NULL, 0, pattern, patlen, &status);
+ if (U_FAILURE(status)) {
+ log_data_err("Fail in udat_open locale %s: %s", locID, u_errorName(status));
+ } else {
+ UChar testString[kUbufMax];
+ int32_t testStrLen = udat_format(testFmt, refDate, testString, kUbufMax, NULL, &status);
+ if (U_FAILURE(status)) {
+ log_err("Fail in udat_format locale %s: %s", locID, u_errorName(status));
+ } else if (testStrLen < 3 || testString[2] != 0x5143) {
+ char bbuf[kBbufMax];
+ u_austrncpy(bbuf, testString, testStrLen);
+ log_err("Formatting year 1 as Gannen, got%s but expected 3rd char to be 0x5143", bbuf);
+ }
+ udat_close(testFmt);
+ }
+ }
+ udatpg_close(dtpgen);
+ }
+}
#endif /* #if !UCONFIG_NO_FORMATTING */
diff --git a/icu4c/source/test/intltest/dtifmtts.cpp b/icu4c/source/test/intltest/dtifmtts.cpp
index a2e7019..4b38a3d 100644
--- a/icu4c/source/test/intltest/dtifmtts.cpp
+++ b/icu4c/source/test/intltest/dtifmtts.cpp
@@ -1062,9 +1062,9 @@
"ja-u-ca-japanese", "H 31 03 15 09:00:00", "H 31 04 15 09:00:00", "GGGGGyMd", "H31/03/15\\uFF5E31/04/15",
- "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GyMMMd", "\\u662D\\u548C64\\u5E741\\u67085\\u65E5\\uFF5E\\u5E73\\u62101\\u5E741\\u670815\\u65E5",
+ "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GyMMMd", "\\u662D\\u548C64\\u5E741\\u67085\\u65E5\\uFF5E\\u5E73\\u6210\\u5143\\u5E741\\u670815\\u65E5",
- "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GGGGGyMd", "S64/1/5\\uFF5EH1/1/15", // The GGGGG/G forces inheritance from a different pattern, no padding
+ "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GGGGGyMd", "S64/1/5\\uFF5EH\\u5143/1/15", // The GGGGG/G forces inheritance from a different pattern, no padding
"ja-u-ca-japanese", "H 31 04 15 09:00:00", JP_ERA_2019_NARROW " 1 05 15 09:00:00", "GGGGGyMd", "H31/4/15\\uFF5E" JP_ERA_2019_NARROW "1/5/15",
diff --git a/icu4c/source/test/intltest/incaltst.cpp b/icu4c/source/test/intltest/incaltst.cpp
index bf0624e..3acb8af 100644
--- a/icu4c/source/test/intltest/incaltst.cpp
+++ b/icu4c/source/test/intltest/incaltst.cpp
@@ -12,6 +12,10 @@
#include "string.h"
#include "unicode/locid.h"
#include "japancal.h"
+#include "unicode/localpointer.h"
+#include "unicode/datefmt.h"
+#include "unicode/smpdtfmt.h"
+#include "unicode/dtptngen.h"
#if !UCONFIG_NO_FORMATTING
@@ -74,9 +78,10 @@
CASE(4,TestBuddhistFormat);
CASE(5,TestJapaneseFormat);
CASE(6,TestJapanese3860);
- CASE(7,TestPersian);
- CASE(8,TestPersianFormat);
- CASE(9,TestTaiwan);
+ CASE(7,TestForceGannenNumbering);
+ CASE(8,TestPersian);
+ CASE(9,TestPersianFormat);
+ CASE(10,TestTaiwan);
default: name = ""; break;
}
}
@@ -636,20 +641,20 @@
// Test parse with missing era (should default to current era, heisei)
// Test parse with incomplete information
logln("Testing parse w/ missing era...");
- SimpleDateFormat *fmt = new SimpleDateFormat(UnicodeString("y.M.d"), Locale("ja_JP@calendar=japanese"), status);
+ SimpleDateFormat *fmt = new SimpleDateFormat(UnicodeString("y/M/d"), Locale("ja_JP@calendar=japanese"), status);
CHECK(status, "creating date format instance");
if(!fmt) {
errln("Couldn't create en_US instance");
} else {
UErrorCode s2 = U_ZERO_ERROR;
cal2->clear();
- UnicodeString samplestr("1.1.9");
+ UnicodeString samplestr("1/5/9");
logln(UnicodeString() + "Test Year: " + samplestr);
aDate = fmt->parse(samplestr, s2);
ParsePosition pp=0;
fmt->parse(samplestr, *cal2, pp);
- CHECK(s2, "parsing the 1.1.9 string");
- logln("*cal2 after 119 parse:");
+ CHECK(s2, "parsing the 1/5/9 string");
+ logln("*cal2 after 159 parse:");
str.remove();
fmt2->format(aDate, str);
logln(UnicodeString() + "as Gregorian Calendar: " + str);
@@ -660,7 +665,7 @@
int32_t expectYear = 1;
int32_t expectEra = JapaneseCalendar::getCurrentEra();
if((gotYear!=1) || (gotEra != expectEra)) {
- errln(UnicodeString("parse "+samplestr+" of 'y.m.d' as Japanese Calendar, expected year ") + expectYear +
+ errln(UnicodeString("parse "+samplestr+" of 'y/M/d' as Japanese Calendar, expected year ") + expectYear +
UnicodeString(" and era ") + expectEra +", but got year " + gotYear + " and era " + gotEra + " (Gregorian:" + str +")");
} else {
logln(UnicodeString() + " year: " + gotYear + ", era: " + gotEra);
@@ -714,8 +719,50 @@
delete fmt2;
}
+void IntlCalendarTest::TestForceGannenNumbering()
+{
+ UErrorCode status;
+ const char* locID = "ja_JP@calendar=japanese";
+ Locale loc(locID);
+ UDate refDate = 600336000000.0; // 1989 Jan 9 Monday = Heisei 1
+ UnicodeString testSkeleton("yMMMd");
+ // Test Gannen year forcing
+ status = U_ZERO_ERROR;
+ LocalPointer<DateFormat> testFmt1(DateFormat::createInstanceForSkeleton(testSkeleton, loc, status));
+ if (U_FAILURE(status)) {
+ dataerrln("Fail in DateFormat::createInstanceForSkeleton locale %s: %s", locID, u_errorName(status));
+ } else {
+ UnicodeString testString1;
+ testString1 = testFmt1->format(refDate, testString1);
+ if (testString1.length() < 3 || testString1.charAt(2) != 0x5143) {
+ errln(UnicodeString("Formatting year 1 as Gannen, got " + testString1 + " but expected 3rd char to be 0x5143"));
+ }
+ }
+ // Test disabling of Gannen year forcing
+ status = U_ZERO_ERROR;
+ LocalPointer<DateTimePatternGenerator> dtpgen(DateTimePatternGenerator::createInstance(loc, status));
+ if (U_FAILURE(status)) {
+ dataerrln("Fail in DateTimePatternGenerator::createInstance locale %s: %s", locID, u_errorName(status));
+ } else {
+ UnicodeString pattern = dtpgen->getBestPattern(testSkeleton, status);
+ if (U_FAILURE(status)) {
+ dataerrln("Fail in DateTimePatternGenerator::getBestPattern locale %s: %s", locID, u_errorName(status));
+ } else {
+ LocalPointer<SimpleDateFormat> testFmt2(new SimpleDateFormat(pattern, UnicodeString(""), loc, status));
+ if (U_FAILURE(status)) {
+ dataerrln("Fail in new SimpleDateFormat locale %s: %s", locID, u_errorName(status));
+ } else {
+ UnicodeString testString2;
+ testString2 = testFmt2->format(refDate, testString2);
+ if (testString2.length() < 3 || testString2.charAt(2) != 0x0031) {
+ errln(UnicodeString("Formatting year 1 with Gannen disabled, got " + testString2 + " but expected 3rd char to be 1"));
+ }
+ }
+ }
+ }
+}
/**
* Verify the Persian Calendar.
diff --git a/icu4c/source/test/intltest/incaltst.h b/icu4c/source/test/intltest/incaltst.h
index 628b6e4..2d42bcc 100644
--- a/icu4c/source/test/intltest/incaltst.h
+++ b/icu4c/source/test/intltest/incaltst.h
@@ -34,6 +34,7 @@
void TestJapanese(void);
void TestJapaneseFormat(void);
void TestJapanese3860(void);
+ void TestForceGannenNumbering(void);
void TestPersian(void);
void TestPersianFormat(void);
diff --git a/icu4j/main/classes/core/src/com/ibm/icu/text/SimpleDateFormat.java b/icu4j/main/classes/core/src/com/ibm/icu/text/SimpleDateFormat.java
index c14c238..737a127 100644
--- a/icu4j/main/classes/core/src/com/ibm/icu/text/SimpleDateFormat.java
+++ b/icu4j/main/classes/core/src/com/ibm/icu/text/SimpleDateFormat.java
@@ -939,6 +939,11 @@
*/
private transient boolean hasSecond;
+ /**
+ * DateFormat pattern contains the Han year character \u5E74=年, => non-numeric E Asian format.
+ */
+ private transient boolean hasHanYearChar;
+
/*
* Capitalization setting, introduced in ICU 50
* Special serialization, see writeObject & readObject below
@@ -1136,11 +1141,20 @@
setLocale(calendar.getLocale(ULocale.VALID_LOCALE ), calendar.getLocale(ULocale.ACTUAL_LOCALE));
initLocalZeroPaddingNumberFormat();
+ parsePattern(); // Need this before initNumberFormatters(), to set hasHanYearChar
+
+ // Simple-minded hack to force Gannen year numbering for ja@calendar=japanese
+ // if format is non-numeric (includes 年) and overrides are not already specified.
+ // This does not update if applyPattern subsequently changes the pattern type.
+ if (override == null && hasHanYearChar &&
+ calendar != null && calendar.getType().equals("japanese") &&
+ locale != null && locale.getLanguage().equals("ja")) {
+ override = "y=jpanyear";
+ }
+
if (override != null) {
initNumberFormatters(locale);
}
-
- parsePattern();
}
/**
@@ -4555,6 +4569,7 @@
private void parsePattern() {
hasMinute = false;
hasSecond = false;
+ hasHanYearChar = false;
boolean inQuote = false;
for (int i = 0; i < pattern.length(); ++i) {
@@ -4562,6 +4577,9 @@
if (ch == '\'') {
inQuote = !inQuote;
}
+ if (ch == '\u5E74') { // don't care whether this is inside quotes
+ hasHanYearChar = true;
+ }
if (!inQuote) {
if (ch == 'm') {
hasMinute = true;
diff --git a/icu4j/main/tests/core/src/com/ibm/icu/dev/test/calendar/JapaneseTest.java b/icu4j/main/tests/core/src/com/ibm/icu/dev/test/calendar/JapaneseTest.java
index 590a847..7b9739a 100644
--- a/icu4j/main/tests/core/src/com/ibm/icu/dev/test/calendar/JapaneseTest.java
+++ b/icu4j/main/tests/core/src/com/ibm/icu/dev/test/calendar/JapaneseTest.java
@@ -19,6 +19,7 @@
import com.ibm.icu.impl.LocaleUtility;
import com.ibm.icu.text.DateFormat;
+import com.ibm.icu.text.DateTimePatternGenerator;
import com.ibm.icu.text.SimpleDateFormat;
import com.ibm.icu.util.Calendar;
import com.ibm.icu.util.JapaneseCalendar;
@@ -150,13 +151,17 @@
@Test
public void Test3860()
{
+ final String jCalShortPattern = "y/M/d"; // Note: just 'y' doesn't work here.
+ final String jCalGannenDate = "1/5/9"; // A date in the above format after the accession date for Heisei era (Heisei year 1 Jan 8)
+ // or the new era in Gregorian 2019 (new era year 1 May 1). If before the accession date,
+ // the year will be in the previous era.
ULocale loc = new ULocale("ja_JP@calendar=japanese");
Calendar cal = new JapaneseCalendar(loc);
DateFormat enjformat = cal.getDateTimeFormat(DateFormat.FULL,DateFormat.FULL,new ULocale("en_JP@calendar=japanese"));
- DateFormat format = cal.getDateTimeFormat(DateFormat.SHORT,DateFormat.SHORT,loc); // SHORT => no jpanyear since we apply "y.M.d" anyway
- ((SimpleDateFormat)format).applyPattern("y.M.d"); // Note: just 'y' doesn't work here.
+ DateFormat format = cal.getDateTimeFormat(DateFormat.SHORT,DateFormat.SHORT,loc); // SHORT => no jpanyear since we will apply a short pattern
+ ((SimpleDateFormat)format).applyPattern(jCalShortPattern);
ParsePosition pos = new ParsePosition(0);
- Date aDate = format.parse("1.5.9", pos); // after accession (Heisei or next). Before accession in the same year would format with the previous era.
+ Date aDate = format.parse(jCalGannenDate, pos);
String inEn = enjformat.format(aDate);
cal.clear();
@@ -207,7 +212,7 @@
// Tests for formats with gannen numbering Gy年
pos.setIndex(0);
- aDate = format.parse("1.5.9", pos); // reset
+ aDate = format.parse(jCalGannenDate, pos); // reset
DateFormat fmtWithGannen = DateFormat.getDateInstance(cal, DateFormat.MEDIUM, loc);
String aString = fmtWithGannen.format(aDate);
if (aString.charAt(2) != '\u5143') { // 元
@@ -228,6 +233,36 @@
}
@Test
+ public void TestForceGannenNumbering() {
+ final String jCalShortPattern = "y/M/d"; // Note: just 'y' doesn't work here.
+ final String jCalGannenDate = "1/5/9"; // A date in the above format after the accession date for Heisei era (Heisei year 1 Jan 8)
+ // or the new era in Gregorian 2019 (new era year 1 May 1). If before the accession date,
+ // the year will be in the previous era.
+ ULocale loc = new ULocale("ja_JP@calendar=japanese");
+ DateFormat refFmt = DateFormat.getDateInstance(DateFormat.SHORT, loc);
+ ((SimpleDateFormat)refFmt).applyPattern(jCalShortPattern);
+ ParsePosition pos = new ParsePosition(0);
+ Date refDate = refFmt.parse(jCalGannenDate, pos);
+ final String testSkeleton = "yMMMd";
+
+ // Test Gannen year forcing
+ DateFormat testFmt1 = DateFormat.getInstanceForSkeleton(testSkeleton, loc);
+ String testString1 = testFmt1.format(refDate);
+ if (testString1.length() < 3 || testString1.charAt(2) != '\u5143') { // 元
+ errln("Formatting year 1 as Gannen, got " + testString1 + " but expected 3rd char to be \u5143");
+ }
+
+ // Test disabling of Gannen year forcing
+ DateTimePatternGenerator dtpgen = DateTimePatternGenerator.getInstance(loc);
+ String pattern = dtpgen.getBestPattern(testSkeleton);
+ SimpleDateFormat testFmt2 = new SimpleDateFormat(pattern, "", loc); // empty override string to disable Gannen year numbering
+ String testString2 = testFmt2.format(refDate);
+ if (testString2.length() < 3 || testString2.charAt(2) != '1') {
+ errln("Formatting year 1 with Gannen disabled, got " + testString2 + " but expected 3rd char to be 1");
+ }
+ }
+
+ @Test
public void Test5345parse() {
// Test parse with incomplete information
DateFormat fmt2= DateFormat.getDateInstance(); //DateFormat.LONG, Locale.US);
diff --git a/icu4j/main/tests/core/src/com/ibm/icu/dev/test/format/DateIntervalFormatTest.java b/icu4j/main/tests/core/src/com/ibm/icu/dev/test/format/DateIntervalFormatTest.java
index 7c7a8a4..8fd76a4 100644
--- a/icu4j/main/tests/core/src/com/ibm/icu/dev/test/format/DateIntervalFormatTest.java
+++ b/icu4j/main/tests/core/src/com/ibm/icu/dev/test/format/DateIntervalFormatTest.java
@@ -710,11 +710,11 @@
"ja-u-ca-japanese", "H 31 03 15 09:00:00", "H 31 04 15 09:00:00", "GGGGGyMd", "H31/03/15\uFF5E31/04/15",
- "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GyMMMd", "\u662D\u548C64\u5E741\u67085\u65E5\uFF5E\u5E73\u62101\u5E741\u670815\u65E5",
+ "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GyMMMd", "\u662D\u548C64\u5E741\u67085\u65E5\uFF5E\u5E73\u6210\u5143\u5E741\u670815\u65E5",
- "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GGGGGyMd", "S64\u5E741\u67085\u65E5\uFF5EH1\u5E741\u670815\u65E5", // The GGGGG/G forces inheritance from a different pattern, non-numeric-only
+ "ja-u-ca-japanese", "S 64 01 05 09:00:00", "H 1 01 15 09:00:00", "GGGGGyMd", "S64\u5E741\u67085\u65E5\uFF5EH\u5143\u5E741\u670815\u65E5", // The GGGGG/G forces inheritance from a different pattern, non-numeric-only
- "ja-u-ca-japanese", "H 31 04 15 09:00:00", DateFormat.JP_ERA_2019_NARROW+" 1 05 15 09:00:00", "GGGGGyMd", "H31\u5E744\u670815\u65E5\uFF5E"+DateFormat.JP_ERA_2019_NARROW+"1\u5E745\u670815\u65E5",
+ "ja-u-ca-japanese", "H 31 04 15 09:00:00", DateFormat.JP_ERA_2019_NARROW+" 1 05 15 09:00:00", "GGGGGyMd", "H31\u5E744\u670815\u65E5\uFF5E"+DateFormat.JP_ERA_2019_NARROW+"\u5143\u5E745\u670815\u65E5",
};
expect(DATA, DATA.length);